lmstudio

LM Studio: download model

download_model

Idempotent

Start downloading a model from the LM Studio catalog or a Hugging Face URL, returning a job_id to track progress. Confirm with the user first, since downloads can be tens of GB.

Instructions

Start downloading a model from the LM Studio catalog (e.g. "qwen/qwen3-4b") or a Hugging Face URL. Downloads can be tens of GB: confirm the model with the user first. Returns a job_id for download_status.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "model": "<model>"
    },
    "name": "download_model"
  }
}

Input Schema

Name	Required	Description	Default
`model`	Yes	Catalog id such as "openai/gpt-oss-20b", or a full Hugging Face URL.
`quantization`	No	Quantization such as "Q4_K_M" (Hugging Face URLs only).

Schema Changelog

Changes observed during successful MCP inspections.

First observedv0.1.0Nov 11, 2026

TDQS

A4.2/5.0

Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare openWorldHint=true and idempotentHint=true, so the network reach and repeat-safety are covered. The description adds genuinely new behavioral context beyond them: this is an asynchronous job (returns job_id, progress tracked elsewhere) and the payload can be tens of GB, which is the operationally important fact. It stops short of stating what happens on re-requesting an already-downloaded model (relevant given idempotentHint=true) or any auth requirement for gated Hugging Face repos.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Three short sentences, front-loaded with source-of-model options, then the size/confirmation risk, then the return contract. No filler and nothing is buried.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

No output schema exists, so the description correctly compensates by naming the returned job_id and pointing to download_status. For a two-parameter, non-destructive operation that is nearly complete; only edge behavior (already-present models, storage/auth failures) is unaddressed.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so both parameters are already documented, including the constraint that quantization applies only to Hugging Face URLs. The description's example catalog id duplicates the schema example rather than adding syntax, naming conventions, or defaulting behavior. Baseline 3 applies when the schema carries the parameter burden.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a concrete verb and resource ("Start downloading a model") plus both accepted inputs (LM Studio catalog id with an example, or a Hugging Face URL). The distinction from the follow-up sibling is explicit: the call returns a job_id consumed by download_status, so an agent can tell this apart from download_status or load_model without opening either schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives a clear precondition ("confirm the model with the user first") tied to download size, and states the sequencing rule with download_status. What is missing is any exclusion or alternative path — e.g. when to prefer list_models to find an id, or how this relates to load_model after the download completes.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.